Audio to Text
Upload or record audio, recognize Chinese and English locally, review segments with a waveform, and export TXT, SRT or VTT.
Checking tool availability…
About this tool
Upload audio or record with your browser microphone. Choose Chinese, English or automatic detection, then listen and edit each segment and its timing before exporting. Review the recognized wording.
How to use
- Upload or record audio and select the language.
- Recognize, listen to segments, and review text and timing.
- Choose TXT, SRT or VTT, then verify and download.
Frequently asked questions
Does auto detection handle mixed Chinese and English?
It can attempt mixed speech, but the model may transcribe some content in the dominant language or miss it. Detected language is an estimate. Review mixed speech and proper names carefully.
What does a review suggestion mean?
It comes from segment-level model cues or missing cues, not an accuracy score. Unmarked segments also need review. Check a flagged segment after reviewing to clear its mark.
How does recording work?
Microphone permission is requested only after Start recording. Pause, resume, stop and use, or discard. Stop and navigation release the microphone. HTTPS or localhost and a supported browser are required.
How long are files kept?
Transcript text stays in private temporary files, outside task options and statistics. Exports inherit source expiry, up to 24 hours, and can be deleted. Select audio again to listen after refreshing or leaving.
Discussion & comments
Share your experience and discover more ways to use this tool.
Loading comments…